On Friday, February 19, Google addressed rumors that they were removing right rail ads with a resounding “yup.” Nothing like a good ol’ Friday announcement from our friends at Google to kick the weekend off, eh?
Obviously this change is going to throw a wrench into how things have worked for the past few years; but change isn’t necessarily a bad thing. To assess the pros and cons of this change, I took a look at trends across about a dozen Elite Paid Search & Shopping clients to assess the potential impact. We’re going to keep monitoring our terms for the next few weeks/months to assess impact post change, but what’s the fun in waiting?
Without further ado, here are my five predictions on the biggest impact this update will have on advertisers, as well as recommendations for what you should do about it.
Warning, long blog post ahead. Cliff’s notes at the end for the lazy.
Prediction 1: There will be a CPC roller coaster.
Top 2 position CPC’s will remain stable. Position 3 -6 will skyrocket and greater than 7 will vanish completely.
In the world of yesterday, the incremental cost for being in prime Top Of Page real estate vs. less desirable Side Rail (right side ads) was significant. CPC increase for Top vs. Side ranged anywhere from a measly 7% increase (for Branded terms in Position One) to a 285% increase (for Non-Branded terms in Position One). Yipes.
It makes sense though; being on the top of the page is more likely to get someone who’s in the conversion mindset. The customer who reads and clicks all the way down the right rail ads is more likely to be price comparing, shopping around or an outright crazy person. The sample shows a significant lift in conversion rate and conversions per impression when ads move from the side to the top. Nice!
For Branded terms, the increase will likely be minimal. Typically you show in position one or two on the top of page anyway; if there are a number of your competitors, chances are they already bid themselves to position 2-3 at the top of the page and they’ll stay that way. Assuming you’re at the top now, you’ll stay at the top. You may feel a bit of upward pressure on CPC’s from below as competitors try to crawl up the page to #4, but unless you have a myriad of re-sellers & affiliates the impact should be minimal.
Non-branded terms are a different story. Get ready for a firefight. Based on the data you can expect to see a 70% lift in conversion rate when non-branded terms move from the side to the top; the difference varies by position, but early data indicates position 4 will see a lift in the area of 3x. Read that again; clicks in position 4 will now likely be THREE TIMES as valuable. Both you and your competitors will adjust accordingly.
Historically speaking, the top of page CPC increase for non-branded terms varied. Moving position one from the side to the top was the most expensive at a 2.9x lift, while moving position two and three led to CPC bumps of less than 2x. Based on the baseline I’d estimate the historic lift to move position four to the top would be about 1.4x side CPC’s.
Problem of course, this isn’t the baseline. Automated bid tools will see this huge bump in conversion rate and bid position 3-4-5 up to the sky, thereby raising the CPC level for anyone. For this change (really any major change), I recommend disabling automated bidding until the dust settles and the new baseline is set.
Prediction 2: Shopping impressions will favor the side vs. the top.
For eCommerce-oriented terms, Shopping and Search have been fighting for the top spot, with little room for error. With no more sidebar text ads, we get more shopping impressions on the right side.
I pulled a separate but similar set of client data from a handful of Elite clients, ranging from eCommerce titans with huge brand recognition to more niche players with less brand recognition. A representative sample if you will. I expected a bit of a shift here and there, but the difference in performance was staggering.
Just as we saw for text ads, there was a pretty massive difference in performance. The surprising part was the difference was flipped; right rail ads CRUSHED top of page shopping ads.
I have the same suspicions with search as I do with shopping ads. Users reviewing shopping ads are price comparing; they’re looking for the most relevant product that looks the best and has the best price. After countless hours of watching my mom use The Google as a navigation tool however, I think I get what’s going on here. When the ads are up top, they stand out dramatically from the rest of search results. The “ooh shiny” factor is leading shoppers to just click on all of the products at the top. Rather than price comparing they’re likely window shopping.
Those who take the time to view the shopping ads on the right rail are more pensive. They take their time to read, they evaluate their choices consciously and they make a cognizant click on the ad that means the most to them.
Based on my data, CTR for sidebar shopping ads is 75% lower than top ads with a slightly higher CPC. However, conversion rate for sidebar ads is a dominant 218% higher, while blended ROAS is 213% higher. I don’t think we’ll see the top-of-page shopping ads disappear any time soon, but I expect volume will shift to the sidebar which is A-OK with me.
I ran a few scenarios based on my data & assumptions. If 10% of impression volume migrates from top to side, our PLA ads should see a 4.5% increase in revenue & ROAS across the same spend. If 20% of impressions migrate, we’d be looking at a 9.5% increase in efficiency assuming everything holds true.
On the contrarian side, we may find that advertisers try to make up for lost volume by increasing aggression in shopping. I suspect we’ll see a larger boost of smaller budget advertisers increasing their CPC’s, which may lead to an overall increase in shopping spend. Even if CPC’s do increase, they’ll come with more profitable sidebar ads which is okay with me.
Prediction 3: True position will matter less. Top of page will matter more.
For this section, I had to revise my dataset since the Brand term data was insignificant below position two. All data is based on non-brand terms only.
Conversion rate does tend to vary between positions 2-3-4, but not as much as you’d might think. CPA remains stable across the top 3 slots despite the differing front end metrics.
Conventional testing wisdom has always said that if you want to try a term out or spread your wings, you could hide out in a lower position to get a baseline for performance. Google’s never been shy about sharing the difference in expected click-through rate, but what conversion rate and CPA has been a challenge to isolate.
What my data found is when on the top of the page, CTR and CPA really doesn’t vary a ton. The sample set showed that the CTR of the top positions were all within about 20% of one another; much less than the drop I’d anticipated. I’ll repeat ye olde broken record that a lot of this is due to behavior. In this case, we’re less worried about navigational behavior and we’re seeing more of users comparing what they see. The logic makes sense; these are typically research queries, so we’d expect that users aren’t just reading ad copy, but also clicking through to find additional value props on the landing pages.
Conversion rate is a different story; it decreased steadily as we left position one before spiking in position four (blame that one on a tiny dataset). First impressions are everything, aren’t they? This was more than made up for by the drop in CPC between positions, so total CPA remained stable.
Now, when the ads are showing on the sidebar in position 1-4 it’s a bit of a different story. CTR decreased as one may expect so total click volume remained the same. CPC steadily increased from position one through four which was a bit of a surprise; while CVR increased in tandem, the ads in position one on the sidebar were dramatically more efficient.
Prediction 4: Position four is going to grow. A lot.
Duh. It’s not just that position four is moving and getting more exposure. I anticipate it will become tremendously more valuable, thereby changing auction dynamics and raising the ad rank floor. Instead of the first page bid setting the base CPC target for a given score, now that target will be at position four. In turn, I anticipate this space will become a lot more competitive thereby raising the prices for positions two and three.
Advertisers would often dip their toes in the water for a set of terms by floating them in position five-six to get a bit of volume and assess the long term value of a keyword. Now, all that testing will have to occur in position four. We’ll have to be more careful with selection of test terms. Going forward it’ll be prudent to test bigger terms in a smaller area, perhaps isolating terms to top performing states as a test. If it doesn’t work in your best location on top of page, it’ll never work.
It’s important to note that as far as we’ve seen (and Google has let on), this change will only affect Desktops. The priority for mobile will continue to be the top two spots, but I anticipate those will grow as well due to this new economic shift. Pay close attention to your mobile modifiers.
Prediction 5: Brand term volume is click volume will switch from organic to paid.
If you want your precious brand traffic, you’re going to have to pay for it. Examples used to be few and far between where you chose not to advertise on your own brand and competitors could take over top of page. CPC’s and low quality scores would wind up being cost prohibitive and competitor ads wound up tucked in the sidebar whether you were there or not.
Going forward, advertising on your own brand terms will be crucial, even more so than they had been. Google and others have done approximately eleventy billion studies showing the incremental value of bidding on brand, so I won’t go into the offensive value, but it’ll be a crucial defensive strategy to make sure you protect yourself from competitors.
Take a look at the screenshots below for example. Blue Apron is facing some competition as well, but their robust brand coverage on the SERP will likely preserve most of the organic clicks. Plated however, is missing a knowledge graph result and facing aggressive bidding from competitors. The organic result is essentially below the fold.
To preserve organic traffic, you’ll need to make sure your brand is covered. Going forward it will be crucial to have core branded terms fire up a knowledge graph, and to have extensive coverage of branded terms (with extensions)
Our best dressed SEO Sr. Manager, Tony Edward, wrote a comprehensive post to explain how you can get a knowledge graph to fire by turning your business (or yourself) into an entity. It’s now required reading.
Bonus Prediction: New ad formats coming!
Action Ads (which were quietly launched and quietly sunset over the past few years), will make a resurgence. The right rail won’t remain naked for long and Google will need to find a way to monetize.
Shopping ads will expand. Currently the right rail shopping format hovers between a 6 and 8 pack of ads, leaving a whole bunch of white space below. Assuming my former prediction comes true, I expect we’ll see an additional row of products added, potentially with position-based bidding for shopping.
Don’t just stand there with your arms crossed pouting – DO SOMETHING. I’m giving the below recommendations to my team and clients, and suggest you do the same for the good of the market.
Turn off any automated bidding for the next few weeks until the dust settles. Your bid tool of choice will see the newfound lift in performance from top ads and bid keywords to the roof. That’s not gonna be good for business. That’s not gonna be good for anyone.
Keep a VERY close eye on mobile bids. While this change is desktop only, chances are a large portion of advertisers will adjust their strategy dramatically without touching mobile. Whatever wonky fluctuations happen over the next few weeks will likely be mirrored in mobile as a result of inattention. Don’t be that guy.
Re-evaluate any test terms and testing strategy going forward. Remember, you won’t be able to “float” in position 5-6 on the first page to test anymore. Isolate Change your test strategy – you won’t be able to float anymore.
Use every extension humanly possible. Now that everyone’s at the top of the page, you better have every clickable link you can to make sure you stand out. Don’t shy away from any upcoming betas either (wink wink).
Lastly, do whatever you can to get a knowledge graph result for key terms if you don’t already. Take owning your brand to a whole other level to maintain organic brand clicks. Or don’t, it’s up to you.
[Too Lazy: Didn’t Read]
This section is for all of you that saw a 2,200 word post, threw up a little and scrolled to the end. I looked at a lot of data and here’s what’s going to happen.
- CPC’s are going to fluctuate. Top of page clicks are roughly twice as valuable as their counterparts on the right rail, but tend to cost three times as much. Turn off bid policies for now.
- As a result of the CPC roller coaster, mobile’s going to show wacky behavior as well. The change is desktop only, but most advertisers who didn’t hire Elite will likely ignore the trees for the forest.
- Shopping impressions are likely to move from top of page to right rail. This is a good thing; top of page has a higher CPC, lower CVR and lower return. Long live the 8-pack!
- Position four is the new first page bid. Expect traffic volume to grow exponentially from this spot, causing a lift in the CPC floor and the disappearance of positions over five.
- Branded organic traffic is going to shift to paid – make sure you have coverage, and get your knowledge graph game on point.
I’ll report back in a month or two with findings after the dust settles. What have you observed with your business?